Maybe you're one of those GMs and you'd like to post your thoughts on (a) why the output of Int/Wis Abillit/Skill action resolution is handled this way, (b) how it affects/propels play, and (c) how it intersects with the type of PC build decision-point thinking cited above?
I like the approach used by Robin Laws in the GUMSHOE role-playing game, where he gives a breakdown of core clues vs. secondary clues (paraphrasing). I've found the critical question is:
"Which information is essential for the players to move forward with the main story they're focused on?" I do not "gate" that essential information behind any check. I just give it to the players when they investigate/reach the appropriate scene.
However, when it comes to non-essential information – monster lore, place lore, faction lore, magic lore, etc. – that's where ability checks (often Intelligence) to recall information come in.
The only exception I can think of to using this approach is if a DM is running a pure sandbox, where there is no "main story" and that, when the players cannot pursue one direction (e.g. due to lack of information), they are expected to flip direction on a dime and pursue another hook. I've never run a pure sandbox, and usually there's a main story the players focus on, so this approach works well for me.
I dislike the lack of stakes in how Intelligence checks are presented. If there are no consequence for a failed check – so there's nothing discouraging pile-on checks or repeating checks – then why roll at all?
Instead, what I do with failed lore checks is use them to either (a) feed misinformation to the players as if it were truth, which requires players who are game to play along, or (b) introduce a complication connecting the PC to the question they're asking (e.g. a failed Intelligence (History) check about a mercenary order might mean that the PC had a run-in with those mercenaries in the past, so there's bad blood; usually I'll introduce this along with a bit of token information but nothing revealing).